When is it best to optimize?

I’m wondering if I should optimize my map entirely at the end, or as soon as I finish a certain section. What do you guys do?

As you progress, otherwise it can be a pain.

Do it as you make it, when you start make it entirly out of nodraw then texture it and then func detail. Thats what I do.

Do it as you go along, build your map with optimization in mind (certainly on the visibility/func_detail/brush count side of things, you’ll be thankful of it. Things like lightmap scales, hint brushes and occluders can be done at the end with no significant issue.


For the nth time, nodraw optimization is almost entirely pointless.

I use dev textures, and then replace them all at the end with nodraw.

Whether it almost be entirely pointless, it does make it easier to work on a map from the outside of the playspace.


You replace all the textures with nodraw?
Your map must look great! :downsrim:

You are hilarious

I texture viewable faces, then replace dev with nodraw. No faces that cannot be seen are rendered.

I didn’t say it doesn’t. I use this exact method so I can hide nodrawn faces and only have to deal with the playable area, but the notion that this classes as optimization irks me and crops up regularly.


VBsp automatically removes faces covered by another or touching the void.

The only situation where significant gain from nodrawing occurs when you nodraw the brushes under a large area of displacements.

I’m not a total mapping noob, but I really have no idea about visleafs or anything like that, can someone give me a beginner’s guide? Luckily I’ve got past the ‘massive skybox box around map’ stage :smiley:

Or brushes that are too high to see with the eye

“significant gain” being the keywords. Assuming your videocard has more than 64mb of RAM you aren’t going to notice the difference on a handful of medium sized rectangular faces



Almost entirely. It’s good practice though so you don’t leave unseen faces rendered (and I mean ones that are still technically visible, but inaccessible to the player.) Even if it’s not a significant gain, a few frames here and there can add up at times. Nothing wrong with not doing it, but it keeps things cleaner and it can’t hurt.

We’re not even talking whole frames if your computer was made post 2005, some people find nodraw hideous to look at when it becomes a huge mass of yellow and it makes it somewhat difficult to make anything out in the 3d view if everything is nodraw, in short it depends entirely on the situation.


Have a clock. I mentioned that early in the thread, I mean if you make your entire map in nodraw first that button is useless because it hides the whole map.

I thought you meant when the map consists of most of it being textured and whatever is left being nodraw areas.

I was insisting that it would be useful at that point, not when you’re constructing with nodraw at the very beginning because that would be plain stupid, you wouldn’t even be able to see what you’re working on.

You cannot really anywya if you just make the whole thing out of nodraw, it just becomes a globular yellow mess, you have to texture as you go along, that’s why some people prefer to work in dev first.